Genuine thinking or random button pressing? self-assessment design in computer-based learning materials

This paper asks how computer-based learning materials can be designed to maximise cognitive engagement on the part of the learner. Computer-based materials are often based around self-assessment questions. How can we ensure that these will stimulate thinking rather than random button pressing? The paper argues that a key issue is that of how feedback is presented, and uses the experience of developing and evaluating a computer-based tutorial based around drag-and-drop style matching questions to discuss four specific questions: At what point do you present feedback? How much information should the feedback contain? How many times do you allow the learner to attempt the question? At what point do you allow users to move to the next question? The paper tentatively concludes that a solution may be to design questions so that it takes more effort for learners to randomly guess the answer than to work out the answer themselves. It is recognised however, that such an approach will depend on the ready availability of the information needed to answer the question, and on the complexity of the subject matter.